Skip to main content

Datacentre operators are still falling short when it comes to keeping track of the environmental impact of their facilities, with the majority failing to keep tabs on the carbon emitted and water used by their facilities.

That’s one of the standout findings of the Uptime Institute’s 12th annual Global data center survey, which is compiled from the responses of 800 datacentre owners and operators, as well as 700 datacentre suppliers, designers and advisers, from around the world.   

The results suggest that just over a third of operators (37% and 39%, respectively) collect and report data pertaining to their site’s carbon emissions and water usage statistics.

When it comes to reporting their Scope 1 and 2 carbon emissions, just 17% said they collected data on the former, while 12% said they did so on the latter.

This is despite 63% of the operators who took part in this year’s survey stating their belief that within five years authorities within the jurisdictions they operate will require them to publicly report their environmental data.

On this point, the survey’s accompanying 33-page report stated: “[Carbon emissions] may quickly become an area of concern for businesses, as most organisations and/or their customers will be required to report this data under new laws, initiatives and rules that are being implemented around the world.”

For this reason, Uptime is advising operators to start reporting all data pertaining to carbon emissions and water usage now before any legal requirements come into force – because this type of data is increasingly being used to decide whether new projects should be given the go-ahead or not.

“A growing number of municipalities will permit datacentre developments only if they are designed for minimal or near-zero direct water consumption,” the report continued. “These types of rules will heavily influence facility design and product choices in the future, mandating cooling equipment that uses water sparingly (or not at all).”

As detailed elsewhere in the report, while the industry’s track record on recording its carbon emissions and water usage is patchy, most operators are much better at collecting data on power efficiency because of the impact this has on the bottom line.

“Most operators collect data that relates to power efficiency, which is as much about saving money as it is about reducing environmental impact,” the report stated.

This is why 85% of respondents said they do report their datacentre’s overall power usage, and 73% said they also – for internal or external purposes – keep a close eye on their facility’s Power Usage Effectiveness (PUE) score.

On that point, the report suggests that PUE scores across the industry are plateauing – and have been for the past several years – after some sizeable declines seen between 2007 and 2018, when the average PUE score shrank from 2.5 to 1.58.

The annual average PUE score now stands at 1.55, which is not – as the report states – because the sector has “reached an efficiency limit” but “largely reflects the broad adoption of inexpensive efficiency measures” that brought about the large drop in PUE scores seen since 2007.

Looking ahead, the think tank predicts that PUE scores could be set to rise in the years to come, despite trending downward over the long-term, as new server processors come to market with high thermal power requirements that might push the limits of air-cooled datacentres.

“If these hotter chips become common within a few years, the industry average PUE may rise before it falls,” said the report.

“Performance requirements and expectations around efficiency will likely push a growing number of operators (and their IT tenants) towards direct liquid cooling (DLC). Greater adoption of DLC could contribute to greater efficiency gains through the 2020s and beyond, both in new builds and retrofits.”

The report’s respondents also spoke of their want to invest in bolstering the resiliency of their datacentres, with 40% stating they have invested in improving the resiliency of the infrastructure housed within their primary sites over the past three to five years.

These investments are paying off, Uptime’s data suggests, with the organisation tracking a “steady improvement in the outage rate per site (or per survey respondent)” with 60% of operators stating they had suffered an outage in the past three years, which is down from 69% in 2021 and 78% in 2020.

Even so, Uptime cautioned against reading too much into these data points, as the effects of the Covid-19 pandemic have made it difficult to draw year-on-year comparisons where datacentre operations are concerned. “It may be too early to call this a strong trend… [and] the level of outages is still high, even if seen to be improving.”  

The report states that when outages do occur, they are becoming an increasingly expensive cost of doing business for operators. “When asked about the cost of their most recent outage, a quarter of respondents say the outage had cost more than $1m in both direct and indirect costs – a significant increase from 2021 and continuing a clear trend. A further 45% say their most recent outage cost between $100,000 and $1m,” the report stated.

The rising cost of outages can be attributed to a combination of factors, the report continued. “Ranging from inflation, fines, service-level agreement breaches and the cost of labour, callouts and replacement parts,” it said. “But the biggest single reason is the growing dependency of corporate economic activity on digital services and on the datacentre. The loss of a critical IT service often translates directly and immediately into disrupted business and lost revenue.”

Andy Lawrence, executive director of research at Uptime Institute Intelligence, said the survey results point to the fact that the sector still has work to do on several fronts to improve how it operates. 

“The global digital infrastructure sector continues to enjoy strong growth and expansion, despite the many obstacles operators are facing today,” said Lawrence.

“We’ve seen the industry invest in increased resiliency and reliability, but there’s still work to be done when it comes to improving efficiency, environmental sustainability, outage prevention, staffing pipelines and more.”

Leave a Reply